Chronic Disease Teaching Tools - Hospital Discharge Data

Hospital discharge data are a rich source of information about the patterns of care, the public health burden, and the cost associated with chronic disease and injury morbidity. These data can be used to gain the attention of policy makers, identify public health priorities, and focus public health programs. The data is also important for conducting epidemiological studies of disease etiology.

New York State maintains a statewide uniform hospital discharge data system, commonly referred to as SPARCS, for reporting hospitalizations.

What do the letters SPARCS (S P A R C S) stand for?

  • SPARCS stands for the Statewide Planning and Research Cooperative System.

When was SPARCS created?

  • SPARCS was established in 1979 as a result of the cooperation between the health care industry and government. In an effort to reflect what is needed by the users of the data, modifications of the data elements and their formats occur periodically. These modifications are a direct result of the input by users of this data.

Why was SPARCS created?

  • SPARCS was originally created to serve as a management tool assisting hospitals and health care organizations/agencies with financial planning and monitoring of patient services and costs. It has since also become a useful tool for the surveillance of disease and injury as well as the conduct of specific epidemiological studies.

How are data collected?

  • Only patients who were admitted to a non-federally regulated hospital are included. Veteran's Affairs and mental health institutions as well as emergency department visits are NOT included in SPARCS. When a patient is discharged from a hospital, a discharge data abstract is created from the information in the patient's chart. NYS regulations require that this abstract be submitted by each acute care facility in the state to the DOH according to a designated format and schedule. Since 1991, hospitals have been required to submit their data electronically. The data submitted is processed and stored in the SPARCS database.

What data are collected?

  • Each year approximately 2.5 million hospitalizations in NYS are reported to SPARCS. The database contains over 120 data elements describing the patient, hospital, health care provider and the hospitalization stay. Patient variables include demographics such as age, gender, race, ethnicity, and residence. Also included is information about the dates and reasons for the hospitalization and what occurred while the patient was hospitalized (clinical diagnoses, procedures performed, etc.). SPARCS also contains information on hospital charges and expected source of payment (Medicaid, Medicare, commercial insurance, HMO, etc.) A unique personal identifier has been recently added to the data set that allows for the linking of multiple records for the same person, and therefore, distinguishes earlier from later hospitalizations for the same condition.

How are the data accessed?

  • Annual data sets are created from the database and are available to the health care industry according to uses prescribed by regulation. The SPARCS data set contains certain elements that are considered "deniable" (i.e. may provide information that would jeopardize the privacy of the patient). To protect patient privacy, release of these elements is subject to special approval of the legislatively established Data Protection Review Board. Approved users sign affidavits attesting that the confidentiality of medical information will be maintained and that data will only be used for approved purposes.

What are the strengths of SPARCS?

  • Because data submission is covered by NYS regulation the database is a complete, centralized, and therefore, a readily available, well-documented source of information on every hospitalization that occurs in NYS. Many specific details about the hospitalization are available, including charge data.

What are some of the limitations of SPARCS?

  • While data are collected monthly from all acute care facilities, annual statewide data sets have been completed approximately 18-24 months after a year has ended. As a result of the new Universal Data Set format, implemented in 1994, which merged multiple data submission formats into one, and a new electronic format, the timeliness of the final data set should be improved to completion within 12 months.
  • The utility of SPARCS data for surveillance activities differs considerably depending on the disease or other condition being monitored. As hospital discharge data, SPARCS represents only those instances of disease and injury that are serious enough to require a hospital stay of at least 24 hours. At the same time, those diseases and conditions that are so severe that they result in death before the patient can be hospitalized, as with some traumatic brain injuries or cardiovascular events, will not be adequately described using SPARCS alone. Further, the SPARCS database counts the number of hospitalizations rather than specific individuals or conditions. Counting the number of hospitalizations for a particular condition may result in inflated numbers because an individual may be hospitalized numerous times for the same condition. Beginning with the 1995 data set, a unique personal identifier is available that will assist with identifying patients who are hospitalized repeatedly, decreasing the chance of counting that individual more than once.

How can I get more information about SPARCS?

  • All aspects of the Statewide Planning and Research Cooperative System, including submission timetables and data requirements are described in detail in the SPARCS regulations. A SPARCS system documentation manual is also available which contains an operations guide, the data dictionary, data submission specifications, and several appendices. Copies of the regulations and system documentation manual can be obtained by contacting the NYS Department of Health SPARCS Administrative Unit at (518) 473-8144.